ElevenLabsTextToSpeech - ComfyUI Built-in Node Documentation

Complete documentation for the ElevenLabsTextToSpeech node in ComfyUI. Learn its inputs, outputs, parameters and usage.

This documentation was AI-generated. If you find any errors or have suggestions for improvement, please feel free to contribute! Edit on GitHub

The ElevenLabs Text to Speech node converts written text into spoken audio using the ElevenLabs API. It allows you to select a specific voice and fine-tune various speech characteristics like stability, speed, and style to generate a customized audio output.

Inputs

Parameter	Data Type	Required	Range	Description
voice	CUSTOM	Yes	N/A	Voice to use for speech synthesis. Connect from Voice Selector or Instant Voice Clone.
text	STRING	Yes	N/A	The text to convert to speech.
stability	FLOAT	No	0.0 - 1.0	Voice stability. Lower values give broader emotional range, higher values produce more consistent but potentially monotonous speech (default: 0.5).
apply_text_normalization	COMBO	No	"auto" "on" "off"	Text normalization mode. ‘auto’ lets the system decide, ‘on’ always applies normalization, ‘off’ skips it.
model	DYNAMICCOMBO	No	"eleven_multilingual_v2" "eleven_v3"	Model to use for text-to-speech. Selecting a model reveals its specific parameters.
language_code	STRING	No	N/A	ISO-639-1 or ISO-639-3 language code (e.g., ‘en’, ‘es’, ‘fra’). Leave empty for automatic detection (default: "").
seed	INT	No	0 - 2147483647	Seed for reproducibility (determinism not guaranteed) (default: 1).
output_format	COMBO	No	"mp3_44100_192" "opus_48000_192"	Audio output format.

Model-Specific Parameters: When the model parameter is set to "eleven_multilingual_v2", the following additional parameters become available:

• speed: Speech speed. 1.0 is normal, <1.0 slower, >1.0 faster (default: 1.0, range: 0.7 - 1.3).
• similarity_boost: Similarity boost. Higher values make the voice more similar to the original (default: 0.75, range: 0.0 - 1.0).
• use_speaker_boost: Boost similarity to the original speaker voice (default: False).
• style: Style exaggeration. Higher values increase stylistic expression but may reduce stability (default: 0.0, range: 0.0 - 0.2).

When the model parameter is set to "eleven_v3", the following additional parameters become available:

• speed: Speech speed. 1.0 is normal, <1.0 slower, >1.0 faster (default: 1.0, range: 0.7 - 1.3).
• similarity_boost: Similarity boost. Higher values make the voice more similar to the original (default: 0.75, range: 0.0 - 1.0).

Outputs

Output Name	Data Type	Description
audio	AUDIO	The generated audio from the text-to-speech conversion.